Intentionally blank page

Advanced Step by step tutorial 3

TARTALOM

1. DYNAMIC ANALYSIS OF A FOUR-STOREY STEEL PLANE FRAME ................................................................................. 5

2. SEISMIC ANALYSIS OF A MULTI STOREY REINFORCED CONCRETE BUILDING USING MODAL RESPONSE
SPECTRUM .............................................................................................................................................................................................. 51
4

Page left intentionally blank

Advanced Step by step tutorial 5

1. DYNAMIC ANALYSIS OF A FOUR-STOREY STEEL PLANE FRAME

Objective The objective of this example is to determine the dynamic response of the following four-storey steel plane
frame considering various dynamic loads.

In this example the capabilities of the AxisVM Dynamic (DYN) module are introduced and a complete time
history analysis is presented. Different types of dynamic loads are applied: two types of support acceleration
(seismic and sine wave acceleration) are used, induced vibration of a rotating (industrial) machine is modelled
by dynamic nodal loads, and the effect of short-term shock load is examined.
Linear elastic behaviour is assumed, however in the analysis the 2nd order P-Δ effect will be considered (geo-
metric nonlinearity).

Suggestions and useful tips are provided for the dynamic analysis using AxisVM. The possibilities for evaluating
the results are briefly described after calculations.
Description of In the task, the creation of the finite element model will not be described. However, the model can be easily
the constructed by applying the knowledge of the previous tutorial examples. The starting file (Four_sto-
finite element
rey_steel_frame_dynamic_analysis_0.axs) is available for download and contains the followings
model

The complete geometry and finite element model of the four-storey steel plane frame. The frame is constructed
in the X-Z work plane.

The columns on the ground floor has a cross-section of HE450 B, while the others have a shape of HE400 B.
The cross-section of the beams above the ground floor is HE400 B, the others are HE320 B.
The cross-sections of the frame are shown in the next figure:
6

The frame is analysed according to the Eurocode standard. The material of the sections is S275, and the ma-
terial is assumed to be linear.

The frame is also assumed to be supported laterally, so only the plane (x-z) behaviour is tested. The nodal
degrees of freedom are set according to this assumption.

The supports of the ground floor columns are fixed (Rx=Ry=Rz=1E+7 kN/m, Rxx=Ryy=Rzz=1E+7
kNm/rad).

The static loads considered in the dynamic analysis have been created and are separated in different load cases
(self-weight, distributed load, concentrated load) as follows:

- The structure is loaded by its own self-weight, which is calculated automatically by the program considering
the set material and different cross-sections.

- On the lower two beams 70 kN/m distributed loads are applied, while on the others 40 kN/m loads are
applied.

- In the middle of the first-floor beams, concentrated loads were placed (in three positions, each with a force
of 20 kN) on a rigid body element at a height of 0.4 m measured from the axis of the beams. These represent
the weight of the rotating industrial machines in the building. These nodes are connected to the beams by
rigid body finite elements.

- In addition, each corner frame node is loaded by concentrated loads (70 and 100 kN) according to the
arrangement shown in the following figure:
Advanced Step by step tutorial 7

self-weight load case:

distributed load:

concentrated load:
8

Start Start AxisVMX4 by double-clicking the AxisVMX4 icon in its installation folder found in the Start – Programs
menu.

Open Click on the Open icon to load the starting file saved on your computer. In the window that appears, choose
the directory containing the file and select it. Click on the Open button to load the model, and the 2D frame
structure will appear in the main window:

First steps Before starting the modelling, check the geometry, the finite elements and the settings described in the image
above.

Load combi- Change to the Loads tab.

nations
Create load combinations for the vibration and dynamic analysis based on the static loads. In our task, the
presented solution is complex, because in the example substantially different dynamic effects are examined in
the same model.

The vibration analysis will be executed in two different ways:

- In the case of seismic, sine wave excitation and shock load, it is sufficient to consider only the horizontal (x)
mass component because the applied dynamic load contains only horizontal components.
- In the case of the machine induced vibration, the vertical component must also be taken into account in
addition to the horizontal one.

The same compilation of static loads will be applied for each dynamic load, but due to the setting of considered
mass components the vibration analysis will be performed on different combinations (the combination name
will be different). The reason for this is that the program binds the results of vibration analysis to the set load
case/combination. If a new analysis based on the same load combination is run with different parameters, the
Advanced Step by step tutorial 9

previous results are not stored. Therefore, two different load combinations will be applied to store all the results
of different analyses.
Note: this trick can also be applied if the effect of other parameters need to be examined and the results of
each analysis have to be saved (e.g. seismic test with different damping parameters).

Click on the Load combinations icon, and the following window pops up:

New row Click on the New row icon, then create two load combinations as follows:
The name of the load combinations should be Co #1 and Co #2. Set the type to ”(user-defined combination)”,
apply combination factors of 1.00 for each load case and leave a reminder comment for each combination –
”for x directional comment” and ”for x, z directional vibration” as follows:

Close the window by clicking OK.

Meshing of Before the dynamic or nonlinear static analysis, the elements must be meshed to get smooth and correct re-
line elements sults. Change to the Mesh tab and click on the Meshing of line elements icon. Select all the beam elements
and set the desired meshing criteria. Divide each element into 4 (equal) segments.

Close the window by clicking on the OK button.

10

By clicking on the Mesh display on/off icon among the Speed buttons, the display of the meshing can be
switched on or off. Now, keep the display of the meshing.

Linear static Change to the Static tab. Perform a linear static analysis to check the load bearing capacity of the frame under
analysis static loads.

Now check the main results and make sure that there is no problem with the finite element model (e.g. insta-
bility).
Note: Always check the model applying simpler load cases or combinations before running time-consuming
nonlinear or dynamic calculations.

In the next steps, the dynamic loads will be defined.

Load groups Change to the Loads tab. Here, click on the Load groups and load cases icon. Beside the existing static load
and load cases, create 4 Dynamic loads. Click on the Dynamic load icon and rename the new load cases as follows:
cases
Name of the load Type of the load to be defined in the load
case case

DYN-EQ seismic acceleration

Dynamic load
DYN-SIN sine wave acceleration

DYN-MACHINE excitation induced by the rotating machines

DYN-SHOCK shock load (short-term concentrated load)

Advanced Step by step tutorial 11

After creating the load cases above, the following can be seen in the window:

Close the window by clicking OK, then specify the loads in each dynamic load case.

Dynamic sup- Firstly, a seismic type excitation will be defined. Select the load case named DYN-EQ (the current load case can
port be checked in the Status palette). As a support acceleration, the dataset named Bucuresti-1986-EW will be
acceleration used, which is included in the database of the software.

Click on the Dynamic support acceleration icon and select the nodal supports of the ground floor columns.

Note: if a command is active, the information about the expected current data can be read in the lower left
blue-grey strip of the main window.

Confirm selection with OK, and the Dynamic support acceleration window will appear.
12

Dynamic load Define the global x-directional seismic acceleration. For the ax [m/s2] component, click on the Dynamic load
library library icon to load the specific function from the database. In the library, select the function named Bucuresti-
1986-EW, and the earthquake diagram will be displayed on the right window. The presented numeric values
are the main minimum and maximum values of the series.

This dataset is the east-west earthquake data series of the Bucharest Earthquake in 1986. It contains data in
0.02 s time steps, and the total duration of the function is 20.42 s. The maximum value is 1.0913. The set is
unitless, it is a so-called load factor function. The support acceleration to be considered in the calculation is
obtained by multiplying this function with a (constant) acceleration value in the given direction. To recover the
values recorded at the the site of earthquake, apply an acceleration value of 1 m/s2. This will be specified in
the next step:
Close the window with OK. In the text box, Set the ax [m/s2] acceleration constant to 1.00 (setting this con-
stant, allows the load factor function to be scaled without modifying the original dataset).

Close the window with OK. In our example, only the global x-directional acceleration is applied (our model is
tested only in the plane of x-z) and the vertical component is neglected. The applied dynamic loads are indi-
cated by the arrows under the supports, and the values above them are the set constant acceleration values.
Advanced Step by step tutorial 13

Note: the user can also specify custom functions using the Function editor. The functions can also be saved
to the database and can be loaded in other models. The use of the editor will be described later.

Dynamic sup- In the second step, define the SINE wave-like support acceleration. Change the load case to DYN-SIN and click
port on the Dynamic support acceleration icon. Select all the nodal supports again.
acceleration

Function Simple functions (sin, exp, sqrt, etc…) can be defined easily in the Function editor using formulas.
editor… Considering the ax [m/s2] component, click on the Function editor… icon.
14

Formula In the Dynamic load function editor window click the Formula icon. In the window that appears, define a
sine wave for the acceleration.

The formula must be given as a function of ”t” (time).

Enter the following formula in the input field of f(t): sin(.5*6.28*t). (Based on the basic formula sin(2*π/T*t)
the period time (T) of the harmonic function will be 2 s, and its frequency is 0.5 Hz.)

The list of the applicable operators can be found in chapter 4.10.28 of the User’s Manual. Using the formula,
the software generates the discrete series considering the given time step (Δt=0.01 s) and the duration of the
excitation (Tmax=10 s). Close the window by pressing OK, and the Dynamic load function editor will appear
again:
Advanced Step by step tutorial 15

On the right side of the window, the specified function can be seen. The table on the left contains the full
discrete data series of the curve considering the given time step. The whole data can be looked through. If
necessary, each data in the set can be edited manually.
Close the editor, and a new window pops up to save the set load function. Save the specified function with
name SIN.

In the Dynamic support acceleration window, specify the acceleration constant value for the load function,
which should be 0.75 m/s2 in this example.

Close the window with OK. The applied dynamic loads are indicated by the arrows under the supports, the
values above them are the set constant acceleration values.
16

Nodal load Next, let’s model the dynamic effect of the rotating industrial machines. We will use a time-varying concen-
trated load to simulate the effect of the rotating machines (rotating around the global y axis). In the example,
it is supposed that the machines have simultaneous phase motion.

Assign the following dynamic loads to nodes 24, 25 and 26.

Advanced Step by step tutorial 17

Change load case to DYN-MACHINE and click on the Nodal load icon and select the nodes mentioned above.
The following window pops up after selection:

The time-varying resultant force is modelled by its global x and z-directional components. For this purpose,
two dynamic load functions should be applied to describe the circular motion. For the x directional force, the
cosine function is used, and in the z direction the sine function is applied.
The time series to be applied starts with a monotonic, linearly increasing part, followed by a pure harmonic
waveform, and ends with a monotonous decreasing part. The maximum amplitude of the sine and cosine
functions is 1.00, the period time is 1.2 s. The complete series was prepared in a Microsoft Excel spreadsheet
because of its complexity. It was generated with a timestep of 0.01 s and the total time is 25 s. This data is in
the enclosed Excel file and the diagrams described above are the following:

Specify the load factor functions of the Fx and Fy nodal forces one after another.
Copy the artificially generated data (time and value columns) into the cells of the Function editor.
Highlight and copy the whole data set starting from the second row. Paste the data into the table in AxisVM
starting in the first cell (t=0 s). Take care to not copy the row of t=0 s, because the first row cannot be
18

overwritten in this way. After completing this procedure, the diagram will immediately appear in the right-hand
window. Compare the result with the curve shown in the Excel table.

Save the new functions with names M_cos (x direction) and M_sin (z direction). Close the editor, set the con-
stant load to 7 kN for both components (the load factor function will be multiplied with this constant value).

Close the window by pressing OK. The applied dynamic nodal loads are indicated by the arrows (in the direc-
tions of x and z) under the supports, the values above them are the set constant load values.
Advanced Step by step tutorial 19

Increase the magnitude of the load in the middle. Click on the sign of the load and change the value of the Fx
and Fz components to 10 kN. After closing the window, the next result can be seen:

Note: a possible modelling method has been presented above. For correct modelling of the effect of a rotating
machine, always ask for detailed information from the manufacturer.

Nodal load Finally, apply a short-term dynamic nodal load as a shock load in nodes 9, 13 and 17 in the DYN-SHOCK load
case.

Change the load case to DYN-SHOCK, and click on the Nodal load icon. Select the nodes marked in the figure
above and confirm selection with the OK. The steps of the data input are similar to before: firstly, the load
factor function should be defined, then the value of the constant load should be set.
20

Function In the line of the Fx component click on the Function editor… icon and define the load factor function
editor… according to the following series of data:

t [s] f(t)
1 0 0.00
2 0.1 1.00
3 1.5 1.00
4 4.0 0.00

The function is defined by the breakpoints. Intermediate values are generated by linear interpolation
based on the time increment set in the dynamic analysis parameters (see later).

Adds a new row The data has to be entered step by step by inserting new data rows and entering the corresponding
values. After entering, we get the next curve as a result:

The maximum is 1.00, and the total load will be scaled by adjusting the Fx constant value. After entering
the data, close the window with OK and save the custom function with name SHOCK.

Additional remark:
If the dynamic load function at the first step (t=0 s) is nonzero, then it acts as a static load on the
structure. In this case the software warns the user before saving the load series:

The Dynamic nodal loads window pops up after saving the function.
Advanced Step by step tutorial 21

Set the Fx constant load to 10 kN and confirm the changes with OK. The given loads are indicated by
the arrows, the value above them are the set constant load values.
Change the value of Fx in nodes 17 and 9. Increase load in node 17. Click on the load that represents
the dynamic load and in the pop-up window, change Fx to 15 kN.
Similarly, reduce the load in node 9, by changing the value of Fx to 5 kN.

The following figure shows the result:

Vibration analysis Change to the Vibration tab and perform a vibration analysis to determine the lowest natural frequen-
cies and mode shapes. This analysis has more roles. In order to avoid a resonance effect, the natural
frequencies must be compared with that of different excitations (sine wave, machine induced vibration).
22

On the other hand, the Rayleigh damping constants in the dynamic analysis will be tuned considering
the dominant vibration shapes in our example.

Click on the Vibration analysis icon, and the next window pops up:

Firstly, the type of vibration analysis has to be selected. The default is the 1st order analysis, but now
let us use a 2nd order analysis.
Note: Second order analysis means that the solution includes the effect of axial forces of truss/beam
elements on the system stiffness. Tension axial forces have a stiffening effect, while the compression
axial forces have a softening effect. These effects influence the free vibrations of the structure.

In the vibration analysis the masses can be converted by the loads defined in a load case/combination
(only the load components parallel with gravity can be converted to masses). As the basis of this proce-
dure, select the correspondent load case or combination. In the list of the available cases, the software
automatically selects the first static load case self-weight. Uncheck this and mark the specific load com-
bination. Two different analysis will be executed based on the mass components to be considered, see
later. Select Co #1 for the first analysis and Co #2 for the second one

Our model does not contain mass (created on the Loads tab), so the program offers the option Convert
loads to masses automatically.

Set the number of mode shapes to 12 in the first analysis and to 30 in the second one. (Considering
vertical mass component, more shapes have to be determined. Beside the main shapes, several shapes
with lower modal mass factor will appears – see the results later.)

Now set the number of Maximum iterations to 30.

As described before, in the first analysis, only the x-directional mass components are needed, the others
have to be switched off. In the second case select also mz component.

Set the parameters as follows, then click OK to start the calculation. After completing the first analysis,
modify the settings according to the following and perform a new analysis.
Advanced Step by step tutorial 23

Settings for the first analysis:

Settings for the second analysis:

24

During the analysis, the user is given information about the progress of the calculation. The Frequencies
panel displays how the frequencies converge step by step. Convergence shows the convergence pro-
cess, but it only shows the convergence of the slowest converging mode shape.

If the calculation is completed, close the window with OK.

Let’s have a look at the determined modes shapes in each case (Co #1 and Co #2). The results of the
analysis are nearly identical considering the x-directional shapes, the main shapes containing vertical
displacement appear only in the higher modes (5-30).

The main data of a displayed shape is also shown in the Status palette.

In our model, only the first five shapes will be dominant based on the two calculations.
Advanced Step by step tutorial 25

1st mode shape (load combination Co #1):

2nd mode shape (load combination Co #1):

26

3rd mode shape (load combination Co #1):

4th mode shape (load combination Co #1):

5th mode shape (load combination Co #2):

Advanced Step by step tutorial 27

Table browser Open the Table browser to see the summarized data of the mode shapes and their modal mass factors:

Based on the parameters set in the Co #1 combination:

In the Co #2 combination:
28

In the list on the left side, select the Frequencies option to display another type of summary table which
contains the frequency, period, angular frequency and relative error of each eigenvalue.

Based on the parameters set in the Co #1 combination:

In a time history analysis, the factors for Rayleigh damping have to be set based on the main, dominant
frequencies. The constants will be tuned by considering a range of frequencies, the boundary values are
the following in our cases:

- Based on the first analysis (Co #1), the lowest frequency is 1.12Hz, the maximum is 9.09Hz,
- In the second analysis (Co #2), the lowest frequency is the same (1.12Hz), and the maximum is
9.20Hz.

Dynamic analysis After the vibration analysis, change to the Dynamic tab.Click on the Dynamic analysis icon and perform
the necessary calculations for the four dynamic load cases considering the following settings:

Load cases:
Advanced Step by step tutorial 29

Simultaneously with the dynamic effect, apply a static load on the frame as well. The first static load
case (self-weight) is automatically selected as default. Change this to the load combination Co #1 for
each run. (Co #2 contains the same loads, it has to be used only in vibration analysis to store the results
of different analysis.)

Note: if no static load case or combination is selected, only the structural mass can be considered or the
given concentrated masses can be converted to loads in the analysis.

Under this option, the specific Dynamic load case has to be selected.

Solution control:
When setting the time increment, the dynamic properties of the excitation and the response of the
structure must be considered. It should be set depending on the highest frequency of the excitation.
Note that increasing the time increment affects the quality and accuracy of the results, the nu-
merical error may increase.
The time increment may differ from the time step specified in the load function, the software uses linear
interpolation to determine intermediate values.

To check the response of the structure, select a typical node in the structure where (greater) displace-
ment is expected under the excitation. During analysis, its displacement, acceleration and velocity can
be tracked. Set the Direction x component (since this is the main direction of the expected motion) and
select the top left node of the frame – node 17.

Linear or nonlinear equilibrium equations are solved with the Newmark-beta method. C matrices are
calculated from the Rayleigh damping constants: 𝐂 = 𝑎𝐌 + 𝑏𝐊, where M are the mass matrices and K
are the stiffness matrices. Constants a and b should be calculated from the damped frequency range
(between fi and fj) and the damping ratio.

The damping function is explained in the next figure and the following formulas are to be used to de-
termine the necessary constant values.

2𝜔𝑖 ∙ 𝜔𝑗
𝑎=𝜉
𝜔𝑖 + 𝜔𝑗
2
𝑏=𝜉
𝜔𝑖 + 𝜔𝑗

where ξ is the damping ratio, ωi and ωj are the angular frequencies related to fi and fj

𝜔𝑖 = 2𝜋 ∙ 𝑓𝑖
𝜔𝑗 = 2𝜋 ∙ 𝑓𝑗 .

For detailed instructions please see chapter 5.3 of the User’s Manual.

Adjusting these parameters strongly depends on the relevant frequencies of the model, the type and
frequency of the excitation and the response of the structure. Due to the nature of the complex curve,
for frequencies greater than fj, or smaller than fi, significantly higher damping can be expected.
30

Therefore, its assignment is primary in the calculation and it requires experience in the field of structural
dynamics.

Without further analysis, apply the next Rayleigh constants in the analysis:

In case of seismic, sine wave and shock load, set frequencies fi and fj related to the 1st and 4th frequency
of the structure equal to fi =1.12 Hz and fj=9.12 Hz. The damping ratio should be 5% (x=0.05). As a
result of the above formulas, set constant as follows: a=0,6265 [1/s] and b=0.0016 s.

Considering machine induced vibration, set frequencies fi and fj related to the 1st and 5th frequency of
the structure equal to fi =1.12 Hz and fj=9.20 Hz). The damping ratio is the same - 5% (x=0.05). Set
the constants a=0.6273 [1/s] and b=0.0015 s.
Advanced Step by step tutorial 31

As an explanation, the function of the damping ratio can be seen in the next figure considering the
results of the 1st vibration analysis:

In the figure, the damping ratios for the first four modes have also been marked. For the 1st and 4th
frequency it is equal to the given ratio, but in case of the 2nd and the 3rd frequencies, lower values are
expected (3-4%).

To Consider static loads and nodal masses in damping, the related function has to be switched on.

Saving results can be controlled: the user can save all the results in each step (Save all steps), or only in
custom intervals (Save at regular intervals).
In preliminary testing of a large or nonlinear model, a greater interval can be specified to save consid-
erable amounts of time. If the model is in the final stage, it can be refined.
Note: the interval of saving does not affect the numeric error. If the interval is increased, less
‘insight’ is expected into the calculated values, so only certain points of the function can be seen.

Check the Convert loads to masses and Follow geometric nonlinearity of beams, trusses, ribs and
shells options.

Set the Maximum number of iterations to 20, the other convergence criteria do not have to be changed.

Perform the analysis according to the above instructions. As an example, consider the following:
32

The Total time [s] should be set for each Dynamic load case. This may exceed the duration of the
excitation (this may be useful if the vibration of the structure must to be tested as well). Exceeding the
total time, the software warns the user before the analysis starts, but it can be ignored:

Our analyses have been completed considering the following time settings:

Load function of the excitation Dynamic analysis

Dynamic load Δt [s] total time [s] time total save results
increment [s] time [s]
DYN-EQ 0.02 20.42 0.02 21.00 save all steps
DYN-SIN 0.01 10.00 0.01 10.00 save all steps
DYN-MACHINE 0.01 24.99 0.05 25.00 save all steps
DYN-SHOCK defined by break- 4.00 0.02 20.00 save all steps
points
In this demonstration example, if the capacity of your computer is limited, the total time can be
limited or the time increment and saving intervals can be controlled. Note that increasing the time
increment can have an influence on the accuracy of the results, the numeric error can increase.

After the required settings, the calculation can be launched by clicking on the OK button. During the
analysis, the software gives information about the current steps of analysis, the results of the tracked
node (aX, vX and eX) and the convergence are displayed on diagrams.

In the following, some examples are presented to show alternative methods for the evaluation of the
results.
Advanced Step by step tutorial 33

Results During the analysis the results are calculated considering the set time interval, but the results that can
be queried depend on the time interval of saving (see before setting the analysis). The results of each
time step are put in separated cases – for example: DYN-EQ [358] (7.14), which contains the result of
the 358th time step (7.14 s) in the DYN-EQ load case.
The results, the internal forces of the model, can be queried in usual way, so the use of basic functions
for result query are not presented in this example. In dynamic tests, the result components are supple-
mented with speed and acceleration data of the nodes.
A specific time step can be selected from the drop-down list by clicking on one of them or the keyboard
arrows can be used to change results step by step. In the last case the displayed results are refreshed in
the background.
Note: each dynamic time step contains the results of combined effect of the set static (here it is the Co
#1) and the dynamic load cases.

In this example, the ordinary methods to inquire results are not described, only a few typical functions
are presented which can be efficiently used in dynamic analysis.

Amination An animation can be made for spectacular presentation of dynamic results.

Select the load case DYN-EQ and the result component ex. Set the scale of the Diagram to 4, and align
the model to the centre. Maximize its scale to fit the window, because the animation only shows the
view that appears in the main window. Click on the Animation icon.
34

The video parameters can be set by clicking on the Setups icon in the lower right corner (e.g. playback
type: unidirectional play / bi-directional play, number of Frames, Frames duration, etc…)

To maximize the resolution of the animation, maximize the size of the Animation window. Generation
of the phases can be started by clicking on the Play icon in the bottom left corner. If the video frames
have been completed, the animation can be saved in AVI video or animated GIF file format. (A presen-
tation video based on the results of load case DYN-EQ is available in the pack of the example.)
With the slider next to the Video file (save) icon, the speed of the animation can be adjusted.

If the result display is also switched on, results are also shown on the animated frames.

Diagram display Creating diagrams can be helpful in evaluating dynamic results. The results can be displayed as a func-
tion of multiple component (result components, time, increment, iteration steps, etc… ), and two types
of graph can be presented in a graph at the same time.

Let us look at some examples of using this feature.

Firstly, check the acceleration of the supports as a function of time in the earthquake type load case. The
result must match the set acceleration function.

From the drop-down list, select load case DYN-EQ [1] (0) (although this is the first load step in the
series, the whole time history diagram will be shown on the diagram).

Click on the Diagram Display icon, and the following window pops up:
Advanced Step by step tutorial 35

By default, the Increment-Time [s] curve can be seen in the diagram.

Diagram Next, set the desired type of function to be displayed. Click on the Diagram parameters icon, and the
parameters following window will appear:

The user can specify two different diagrams to be shown in one graph (the result components on the x
and y axes can be also different).
First, let’s display a single curve, the acceleration curve of node 1 (support) as a function of time. Take
the time on the horizontal axis, select Time [s] from the drop-down list of Component X1. At the list of
the Y1 component, look for aX [m/s2], which is the horizontal, global X-directional acceleration. Click
on the Node button, and in the main window select node 1.
36

Confirm changes by clicking on OK.

The following will appear in the Diagram display window:

Under the graph, there is a scale which represents the Time axis and it contains green and black markers
at the ends. Moving the black marker along the axis, the current result value will be shown on the dia-
gram according to its position. Moving the green ones allows the range of the displayed function to be
set. The next figure shows the effect of moving the markers.
Advanced Step by step tutorial 37

Fit in view in X di- If the range of the diagram has been limited by the green markers, then the graph can be stretched by
rection clicking on the Fit in view in X direction icon.

Table The results of the calculated time steps can be presented in a table. Click on the Table icon, and a table
appears next to the graph on the left side, which contains the whole data. If it is necessary, these cells
can be copied to the clipboard for further processing (using Excel, or saving the data file, etc…).
38

By clicking on a cell in the table, the result label jumps to the selected time step, then the black marker
on the bottom axis turns grey (it will be active again when moving this marker to a specific point).

Diagram Now, let’s show the My flexural moment of beam no. 27 in node 5 as a function of time. For reference,
parameters the specific beam and its end node are marked in the next figure.

Click on the Diagram parameters icon.

On the x1-y1 diagram, select the Time [s] for the X1 component, and select the My [kNm] result
component for Y1. Next, select the correspondent node (5) and beam (27) by following the instructions
below:
Click on the Element button, then the main window becomes active.
With one click select node 5, then one of the beams which belongs to the selected node will be high-
lighted automatically in purple. The following status will be shown on the screen:
Advanced Step by step tutorial 39

The software has automatically selected beam 1 (the beam is divided into finite elements, the program
highlighted only the upper finite element), however we need to select beam 27. By pressing the Tab key
on the keyboard, one can switch between the beams (that have the same end node). Press the Tab key
repeatedly to select the correct element (in our case just press it once to highlight beam 27). As a result,
the following selection is obtained:

Click on node 5 again to close the selection, and the Diagram display parameters window will appear
again:
40

Click OK to accept the changes, and the required curve is presented in the Diagram display window.
Drag the green markers to the ends (0-21 s) to display the whole diagram.

Diagram Show the displacement of node 1 (lower, support node) and node 17 (top level) as a function of time
parameters on the same diagram to see the relative displacement between nodes (or levels).
Click on the Diagram parameters icon.
On the x1-y1 diagram select the Time [s] for the X1 component, and for Y1 select the eX [mm] result
component and select node 1.
Activate the x2-y2 diagram with its checkbox. At the X2 component the default setting Time [s] is
suitable for our purposes, but for the Y2 component select the eX [mm] component. Finally, select node
17.
Finally, the next result can be seen in the Diagram display parameters window:
Advanced Step by step tutorial 41

Click on OK to close the window.

In the Diagram display window turn off the Table, if necessary drag the green markers to the end
positions (if only a range of the diagram is visible) to see the whole diagram:

The blue curve shows the displacement of node 1, the red curve belongs to node 17. The colour of the
curves and their labels give information about the type of the curve.
Same range on the It can be seen on the figure that the program used different scales on the vertical axis (see the blue and
two Y axes red vertical axis). The same scale can be set by clicking on the Same range on the two Y axes icon.
42

On the bottom axis, move the black marker to a desired position to show the displacements values of
the nodes. The relative displacement can be calculated by taking the difference between the results.
If the Table function is opened, then the whole data of the diagrams can be easily read and compared.
If it is necessary, the data can be copied and evaluated in any other spreadsheet program (e.g. Excel).
For more details about the using of the special features of the current window, please see the User’s
Manual. Close the Diagram display window by clicking on the OK button.

Diagram display Next, follow the motion of one of the excited nodes in the DYN-MACHINE load case. In the main win-
dow change the load case to the DYN-MACHINE [1] (0).
Click on the Diagram display icon, and in the window that appears click on the Diagram parameters
Diagram icon.
parameters Only the x1-y1 diagram is needed, therefore uncheck x2-y2 diagram (and their parameters will be-
come disabled). By finishing the settings, the following can be seen in the parameters window:

Click OK to see the result diagram:

Advanced Step by step tutorial 43

Due to the excitation and movement of the structure, a quasi-steady movement similar to an ellipse is
formed. Close the Diagram display window by clicking OK.

Diagram Analyse the displacement of node 17 in a diagram.

parameters
Click on the Diagram display parameters icon and set the following:

Close the window with OK.

44

The following result can be seen displaying the whole range of the diagram:

Limit the range of the presented curve to see two internal waves as shown below. Adjust the markers to
the following limits: 5.35 - 8.00 s.

In this example, a quasi-harmonic motion is induced by the excitation. If the time increment or the den-
sity of the saved steps is insufficient, then the curve will be ‘piecewise’, and the accuracy of solution will
not be adequate.

Take a look back, the load function was defined in 0.01 s time steps, but the calculation was performed
and results were saved considering 0.05 s time steps. The appropriateness of the applied time step can
be checked by analysing the result curves by counting the intermediate points of a ‘wave’. (It is our
experience that a clear harmonic wave - e.g. sine / cosine - must contain at least 10 intermediate points
to obtain acceptable results. For more complex / precipitous curves/excitations, it is worth testing the
effect of different time intervals to find an adequate setting).

Below is a result of a deliberately wrong adjustment. The calculation was performed with a time incre-
ment of 0.2 s. There is significant error in the minimum and maximum values, and the shape of the
curves clearly indicates it.
Advanced Step by step tutorial 45

Diagram display
Finally, let us analyse the damping effect of the frame examining in the DYN-SHOCK load case (the
duration of the shock effect is 4 s, but the total time of the performed analysis is 20 s). Select the DYN-
SHOCK [1] (0) load case and return to the Diagram display parameters window. Only the x1-y1 dia-
gram has to be used, the other has to be switched off. Set the displacement of node 17 to display as a
function of time as follow:

Uncheck the Markers option, but the Grid should be displayed. Close the window with OK, and the
following will be the result diagram:
46

In the figure above, the orange vertical line (it was not created in AxisVM) indicates the time when the
dynamic nodal load becomes zero and from where the free vibration of the frame starts. The calm down
tendency of the waves points to the set damping behaviour of the structure.
Result In many cases, it is needed to evaluate the results of various dynamic loads together using envelopes.
display Present Sminmax results are an envelope of the results of DYN-MACHINE and DYN-SHOCK load cases.
parameters From the drop-down list of results, select Sminmax under the Beam stresses components.
Then, a new custom envelope has to be created. Click on the Result display parameters icon, and the
Result display window pops up:
Advanced Step by step tutorial 47

In the top line, click on the Envelope function.

New Click on the New envelope set icon above the left white field. As a result, a new envelope set will be
envelope set created, which appears in the list under the icon. Click on its name and rename it as ‘Custom’.

By default, the new envelope will contain all available load groups and cases. Modify the set by the
following: in the list named Envelope, only the DYN-MACHINE and DYN-SHOCK load groups must be
kept as checked, the others have to be unchecked. If it is necessary, the load groups can be folded by
clicking on the ‘–‘ sign before the name of the group to hide their time steps.
Finally, change the Display mode to Filled Diagram.
48

Confirm results by clicking on OK.

Note: each time step in the dynamic load groups can be checked or unchecked individually when an
envelope set is created. As an example, see the following figure:

In the main window, find the new envelope – Custom - in the drop-down list and select the display of
the Envelope Min, Max results.

By querying any result component, the envelope result of the set load groups / load cases can be seen
in the window.
Advanced Step by step tutorial 49

Cross section Display the Cross-section beam stresses (Sminmax, Vminmax, Sominmax) in beam 27 by selecting
stresses the DYN-EQ load case. In the main window select the DYN-EQ [1] (0) load case and click once on the
beams, then the following result window will appear:

The figures in the left hand shows the stress results along the axis of the beam considering the current
time step.
In the middle lane, the different stress diagrams of the selected cross-section and their main values are
shown. In the table displayed in the upper right corner, the stresses in the calculated points can be read
(these points are marked on the cross section under the table, their position can be edited or modified
in the cross-section editor). Presently, the results of the cross section in the half of the beam are illus-
trated.
50

The stress results of any cross section along the beam can be queried by dragging the vertical axis of
the figures to the desired position. In the lower right corner, a schematic axis can be seen which also
shows the actual position (x[m]=2.500), the length of the element and the node number of the end-
points. The current position can be controlled by overwriting the current value or dragging the vertical
axis to the right position. The latter function is more convenient for taking a specific position.

Display the result of the cross-section at position 4.50 m (measured from the startng point – node 5 –
of the beam). Overwrite the position value in the lower right window. The following result can be seen
in the window after adjustment:

The user can switch between the results of each time step in the drop-down list by clicking on a specific
step or by navigating up or down with the arrows keys. In the latter case, the results will automatically
be updated in the background:
Advanced Step by step tutorial 51

2. SEISMIC ANALYSIS OF A MULTI STOREY REINFORCED CONCRETE

BUILDING USING MODAL RESPONSE SPECTRUM
Objective Perform the seismic test of the following reinforced concrete (RC) building using modal response spectrum
analysis. The basic finite element model file of the building is enclosed to this manual (MRSA_of_RC_build-
ing_0.axs).

The earthquake test is carried out according to the basic EN 1998-1:2008 independently of any national
annex (in real projects the user has the choice to select that code which contains the specific rules for the
current country). In our task, the so-called Modal Response Spectrum Analysis (MRSA) is used: the earthquake
loads are generated according to the standard based on the dominant modal shapes of the structure. The
distribution of the seismic load is calculated by the magnitude of the oscillation and the distribution of the
masses in each modal shape. After, the seismic results of each modal shape are determined by a linear static
analysis. Finally, the (critical) design results (internal forces and displacements) are obtained by summing up
the results of the shapes using standard methods (e.g. SRSS, CQC).
Note: this kind of seismic analysis supposes linear behaviour and uses the principle of superposition, therefore
nonlinear material or finite elements cannot be considered in MRSA.

Based on the seismic characteristics of the location of the building, a reference peak ground acceleration
0.10 g has to be considered applying type I response spectrum and soil class C. Take the behaviour factor
to 1.50.

Description of In the task the creation of the basic finite element model will not be described, the model can be easily built
the finite ele- up applying the knowledge of our previous examples. Primarily, the main steps of applying MRSA method
ment model and
will be detailed using the available tools of the program. The starting file (MRSA_of_RC_building_0.axs) is
loads
enclosed to the example, which contains the followings:

The full geometric and finite element model of the multi storey, reinforced concrete (RC) building without the
generated seismic loads and results.

The building has a ground floor and 5 storeys, and its main function is office. The characteristic structural
dimensions of the building are the following: the overall plan size is 23 x 22 m and the total height is 22.5 m.
The main structure is built as monolithic reinforced concrete structure (monolithic walls, columns and flat
slab), but above the 4th floor there is a steel-structured mechanical unit.
The rigidity of the reinforced concrete structure is ensured by RC walls, while the steel structure on the roof
is braced by steel struts.
52

The thickness of the RC walls is 200 mm, the flat slabs are 220 mm. The section of the RC columns varies
according to the internal forces. The intermediate columns are made with circle shapes, while the columns
behind the facades have a rectangular cross section.
The columns of the steel structure are HE 160A, the beams are made of HE 200A. The bracing elements are:
N80x80x4,5 mm.

The structural roof of the installation unit is not modelled in detail, it is assumed that works as a rigid plate in
the horizontal plane. According to this assumption, the roof is modelled with a rigid diaphragm. The loads of
the roof are placed on load panels which transfer the loads to the structural elements above it.
Note: the structural role of the top steel roof is subordinate in the main behaviour of the entire building,
which makes it possible to make simplifications and to present the using of rigid diaphragm element in our
example. Simplifications can lead to a more perspicuous model, but in all cases the user should consider all
the assumptions.

In the model, the basic design code called Eurocode is applied (see the Design codes… window in AxisVM).
The RC structures is made of concrete with a strength of C30/37, the material quality of the steel profiles is
S235 uniformly.

Setting the nodal degrees of freedom, every node is considered as free.

The supports on the ground floor are fixed, their stiffnesses are the following: Rx=Ry=Rz=1E+6 kN/m, and
Rxx=Ryy =7.5E+6 kNm/rad (using global coordinate system). The stiffnesses of the walls are defined relative
to the edge using the next values: Rx=Ry=Rz=5E+5 kN/m/m and Rxx= 5E+5 kNm/rad/m.
Note: the result of the earthquake design is significantly affected by the stiffness of the supports, but this
effect is not examined in our example. The stiffness of the supports can be estimated based on the soil physics
characteristics assuming a dynamic effect (the stiffness loads may be different in case of static loads).

For the analysis, the main load cases have already been created and been separated into load groups. The
given loads are the so-called characteristic values. These values are not specifically described here, please
have look at the model.
In earthquake analysis, it is not necessary to adapt partial loading. Usually, the total load has to be considered
using the specified seismic combination factors (probably the total load will be the critical, hence this provides
the greatest mass). The model does not contain meteorological loads (wind and snow) because they are not
simultaneous (y2=0) with the seismic load.

For the vibration analysis a custom load combination is still needed which It is not included in the starting
model. This combination will be defined later.

The seismic loads and cases will be automatically generated based on the set basic earthquake parameters
and the main modal shapes considered.

The critical minimum and maximum results of the total analysis will be determined automatically considering
the set load groups and combination factors, therefore no need to define any more custom load combinations
for this purpose.

Parts in the For the better visibility and manageability, parts have been defined considering structural types (walls, col-
model umns, slabs related to different storeys) in the model, which have been sorted into various directories as seen
below:
Advanced Step by step tutorial 53

Tricks If parts are used in displaying and the function Perform design calculations for visible parts only is checked
in the menu of Settings/Preferences/Parts, then the derivative results will only be calculated (e.g. calculation
of necessary reinforcement in RC elements) for the visible elements, not for the entire model. This option can
save time in processing the results of large models, where the complete calculation of the derivative results may
require a substantial amount of time.

Start Start AxisVMX4 by double-clicking the AxisVMX4 icon in its installation folder found in the Start – Programs
menu.

Open Click on the Open icon to load the starting file saved on your computer. In the window that appears, choose
the directory containing the file and select it (MRSA_of_RC_building_0.axs). Click on the Open icon to load
the model. After, the model of the building appears in the main window:
54

First steps Save the model under a different name (if necessary, we can return to the original model later). In the File
menu, look for the Save as… function and click on it, then rename the model: MRSA_of_RC_building_0.axs.

Before starting the seismic analysis, check the model, its geometry, the created finite elements and the settings
listed above.

Stories In the first steps let us define stories (main levels of the building), which can have several roles during the
analysis. Stories can be displayed separately, they behave like the Parts and give help in constructing the
model and ensures better visibility.
Stories have also a role in defining the torsional effect of the seismic action because the eccentric forces can
be assigned to these levels. The latter function will be presented later.
For more benefits and opportunities of this feature please see the User’s Manual.

Clicking on the Stories icon on the left-hand vertical toolbar, the following window pops up:
Advanced Step by step tutorial 55

Find Stories can be specified individually, entering the height value of each storey, but auto search function can
also be used. Click on the Find icon, then the function will search for the horizontal slabs with different levels.
The next will be the result:

Add The level of the steel roof (+22.50 m) has not been recognized by the program, because the roof was mod-
elled by diaphragm, not with slab. Add this level to the list as well. Enter the height data (22.50 m) in the Z
[m] = input field, and then click on the Add icon:
56

The generation of stories has been completed with this final step, click on OK to close the form.

Load In MRSA the seismic load is defined based on the mode shapes. Therefore, a vibration analysis has to be done
combinations using a load combination, which follows the seismic combination rules specified in the standard (the masses
for the vibration analysis are obtained by converting the loads of that combination). In the software, this
combination should be defined as ’custom’. Complete the followings to create this combination.

Change tab to the Loads and click on the Load combinations icon. In the appearing window click on the
New row
New row icon, then the software creates a new row in the database. Enter the following name into the first
cell: SEISMIC. Set the type of the combination to „- (user-defined combination)”, finally specify the combi-
nation factors to the load cases according to general formula:

∑𝑗≥1 𝐺𝑘,𝑗 " +" ∑𝑖≥1 𝜑 ∙ Ψ2,𝑖 ∙ 𝑄𝑘,𝑖 .

Permanent loads take part in the combination with their characteristic value. Variable loads must be consid-
ered by their quasi-permanent value. The factor of f is usually 1.00 (for the benefit of security).
Note: this combination is only valid for the vibration analysis which precedes the generation of seismic loads.
Therefore, it contains only permanent and variable actions, the accidental seismic load is not included in it.

The following table summarizes the combination factors to be entered:

Name of the load case combination factor

self-weight 1.00
finishes 1.00
susp-ceiling 1.00
facade 1.00
staircase-dead-load 1.00
1st-floor-office 0.30
2nd-floor-office 0.30
3rd-floor-office 0.30
4th-floor-office 0.30
1st-floor-balcony 0.30
2nd-floor-balcony 0.30
Advanced Step by step tutorial 57

3rd-floor-balcony 0.30
4th-floor-terrace 0.30
5th-floor-flat-roof 0.30
1st-floor-div-wall 1.00
2nd-floor-div-wall 1.00
3rd-floor- div-wall 1.00
4th-floor- div-wall 1.00
5th-floor-installation 1.00
6th-floor-steel-roof 0.00
staircase-var-load 0.30

Entering the data above, confirm changing and close the window with clicking on the OK button.

Domain mesh- Before the dynamic (and nonlinear) analysis, domains and line elements must be meshed. In our task, it is
ing sufficient to divide only the domains, not necessary to mesh the columns, because the significant part of the
total mass is concentrated on the stories and the bending mode shapes of the columns are not dominant in
the behaviour of the entire structure.
Note: if a line element is not meshed, then the self-weight or any loads on the element (which will be con-
verted to mass) are distributed to the end nodes in the vibration analysis.

Depending on the size and type of the building, various size and density of mesh can be applied.
In earthquake analysis, it may be appropriate to provide the vertical stiffening (wall) structures with a dense
mesh, while less dense mesh can be used in the slabs.
The dynamic behaviour of the structure is primarily determined by the stiffness and behaviour of bracing wall
systems not by that of the slabs. In many cases, the slab can be regarded as rigid diaphragms if its stiffness is
great enough (see the relevant parts of the standard) and the vertical effect of the earthquake is negligible
compared to the static loads.
Optimization of the applied mesh size can reduce the number of finite elements, which can save time in
computing and evaluating the results.

In our model, we only use a single mesh because applying various mesh size does not result significant
savings due to the small size of the building.

Change tab to the Mesh and click on the Domain meshing icon and select the entire model with * (All) icon.
58

In the form, select the Triangle mesh, use Average mesh element size - 0,8 m, check the Adjust mesh to
column heads (to enable cutting of moment peaks) and apply Uniform mesh size. Start the meshing pro-
cedure by clicking on the OK button.

When the meshing procedure is finished, check the regularity of the mesh while rotating the model.
Mesh Among the Speed buttons the display of mesh can be turned on or off by clicking on the Mesh display
display on/off on/off icon.
Advanced Step by step tutorial 59

Stiffness In case of RC structures, the effect of cracking can be considered by using the Stiffness reduction… option
reduction… (EN 1998-1:2008 recommends 50% stiffness reduction to RC elements). This reduction is considered in the
vibration and static analysis as well.

The Stiffness reduction for response spectrum analysis form can be found in the Settings menu or it can
be set among the Vibration analysis parameters. Adjust the reduction parameters now, before changing tab
to the Vibration. Click on the Settings menu and in the list find and activate Stiffness reduction… function.
As a result, the following window pops up:

In the form, the basic stiffness reduction parameters can be set for the Displayed parts, the Entire model or
the Selected elements. If the entire building is displayed in the main window and there are no selected items,
then the Displayed parts and Selected elements checkboxes are inactive, as shown in the figure above.

In case of Columns, Beams and Other elements (beam), two options are offered: the compressive and tensile
stiffness (A) and the inertia (I) can also be reduced independently from each other.
For Walls, Slabs and Other domains, only one parameter can be specified, which reduce also the bending,
shear and membrane stiffness of the finite elements.
The numbers in the parentheses next to the finite element type show how many elements are in the entire
model/in the displayed parts/or among the highlighted elements.
Consider the following settings shown in the figure: reduce the stiffness to 50 percent (0.5) for each element,
but in case of Columns, Beams and Other elements only the bending stiffness (kI) should be lowered.

Finishing the data input, confirm the changing and close the window with clicking on the OK button.
60

Vibration Seismic load generation requires information on undamped free vibration frequencies and corresponding
mode shapes. Thus, the first step in creating seismic loads is the calculation of a sufficiently large number of
mode shapes and corresponding frequencies of vibration.

Change tab to the Vibration and complete an analysis according to the following.

Click on the Vibration analysis icon.

In the window that appears, set the required parameters. Let us go through the options:

At the top left of the window, the type of the analysis can be selected. By default, Vibration (1st order) is
selected, which is appropriate for our investigation.
Note: In 2nd order vibration analysis the solution includes the effect of axial forces of truss/beam elements on
the system stiffness. Tension axial forces have a stiffening effect, while the compression axial forces have a
softening effect. These effects influence the free vibrations of the structure.

In the list under the type of the analysis, the load case or load combination can be selected as the basis of the
vibration test. The program automatically selects the first load case (self-weight). Uncheck this load case.
Advanced Step by step tutorial 61

After find and check our custom load combination called SEISMIC at the end of the list, under Load com-
binations.
Note: at once, user can select more load case/load combinations. In this case, the program performs as many
independent calculations as the number of cases have been marked.

In vibration analysis, if masses are not specificially defined, masses can be determined based on the selected
load combinations or load cases by converting them to masses. In our case the program automatically offers
the Convert loads to masses option, since it does not contain masses.

Set the Number of mode shapes to 20. This number depends heavily on the complexity of the model, the
size of the building and the number of stories. (In our task, 20 mode shapes will be sufficient to reach at least
90% of the total mass of the structure in the directions considered.)

Set the number of Maximum iterations to 30. If the model does not converge properly, we may need to
change it. The Eigenvalue and Eigenvector convergence should be left as default.

Diaphragm: when running a vibration analysis with checking the option of Convert slabs to diaphragms, all
slabs (horizontal plates) will be temporarily replaced by diaphragms (always check whether the conditions
described in the standard are satisfied).
The running time may be reduced if the model contains only columns and slabs. If structural walls are included,
the number of equations will be reduced but the bandwidth will be increased. The resultant running time may
be greater than without diaphragms.
If there are ribs or supports in the model that are connected only to slab to be converted to diaphragm, this
function cannot be used. If this problem occurs, the program gives an error message at the beginning of the
calculation:

Note: the slabs/base plates which are supported (e.g. with surface support) will not be converted to dia-
phragm even if this function is selected.

Stiffness reduction for response spectrum analysis: here, the stiffness reduction of RC elements can be
specified Previously, the relevant parameters have been set, therefore this setting does not have to be done
again, but mark the checkbox to apply Reduced stiffness.

Use increased support stiffness:

Supports act in a different way during vibrations. Using increased support stiffness (1010 kN/m for nodal
supports, 107 kN/m/m for line supports, 104 kN/m2 for surface supports) may help to get more realistic
vibration results. Do not check this option now, the stiffness parameters of the supports have been declared
for seismic an dynamic actions.

Masses: in the calculation, only the x and y mass components should be considered (the vertical effect of
the earthquake can be negligible depending on the country’s regulations).

Configuration of Masses considered: beside All masses it is possible to reduce the masses to those above a
given Z height or a given level (if the model contains levels). An example could be the exclusion of the base-
ment from the vibration analysis, the self-weight and other types of loads are not considered below the
ground level, including the base plate (it the model contains base plate).
At this option the default setting - All masses - has to be selected.

For more details about the parameters and the settings above, see the relevant sections in the User’s Manual.

Set the parameters shown below as a summary, then click OK to start the analysis.
62

During the calculation, the frequencies of mode shapes and the slowest convergent shapes can be traced
as a function of the iterations.
Advanced Step by step tutorial 63

The analysis stopped at 23th iteration step because the set convergence criteria have been met, no need to
continue until the final 30th step.

Click on OK to close the window.

Always check the resulting mode shapes. If there is an error in the model, analyzing the mode shapes may
help to filter out the main problems (instability, any problem in stiffness). On the other hand, we get an over-
view about the dynamic behaviour of the building (e.g. which are the main shapes, our structure is sensitive
to the torsional effects or not, etc…).

The first three shapes are shown below displaying the displaced shape applying a zoom scale of 10:

1st mode shape:

64

2nd mode shape:

Advanced Step by step tutorial 65

3rd mode shape:

The Status palette on the left display information about the mode shapes that can be helpful in the
evaluating of results.

Let us go through these, select the 1st mode shape:

Under the title of Vibration analysis, it is indicated that the calculation was actually performed with
using Stiffness reduction (see the explanation for this function earlier).

Below the applied design Code is shown.

The Case shows the load case/ load combination applied in the vibration analysis (now it is SEISMIC).

Mode shows the ordinal number of the mode shape (the first from the calculated 20 shapes).

In the next unit of the palette, the main results of the actual mode shape are presented (frequency, period,
circular frequency, eigenvalue, error). The list ends with the maximum number of iterations (22): the cal-
culation has been finished at the iteration step of 22th.

The unit of Modal mass factors provides information about the followings:
- the modal mass factor in each direction (eX és eY) which belongs to the displayed mode shape
(if the global Z component of the mass is taken into account in the analysis, it also appears in
the list).
66

- Status informs about the active or inactive state of the displayed mode shape – see the expla-
nation later.
- under the Status, the total modal mass factors achieved by all the shapes set as active. Now,
these pass the limit of 90% in both calculated directions x and y, so the determined number of
shapes (20) are sufficient. No need to calculate more shapes.
- Comp. shows the actual result component that is being displayed (eX).

Table browser The results of the vibration analysis are presented in the table of Modal mass factors which is available
in the Table browser. Click on the Table browser icon.

It is extremely important to meet the standard requirements which refer to the total mass factor. The
general EN 1998-1:2008 has two requirements:
(1) the total considered modal mass in each direction shall exceed 90% of the total mass;
(2) all mode shapes corresponding to a modal mass larger than 5% of the total mass shall be
taken into consideration.

The sums of modal mass factors can be checked at the bottom of the table.

The rows of the table show the main results of the determined mode shapes and the corresponding mass
factors. The columns named Si eX ; Si eY and Si ez show the summed mass factor in the main directions,
but only the results of shapes set as active are summed up.

Among the shapes, there are many shapes which mass factor is negligible (e.g. less than 0.01). The neg-
ligible shapes can be switched to inactive, but make sure that the total mass factor in each direction does
not go under the limit of 0.90 (90%). The inactive shapes will be ignored when generating the seismic
load. (Keep in mind that neglecting some of the shapes also proportionally reduces the total seismic
load.)

Every mode shape that has been indicated as active will result one or two (even if the torsional effect is
considered) seismic load case per direction. Thus, the filtering of the modes shapes is recommended if
there are plenty of shapes, as it can significantly reduce the number of load cases and the required cal-
culation time as well.

The active status of the mode shapes can be switched in two ways. We can go through on each row and
at the end of the rows the status of the current shape can be modified by clicking on its field. Or standing
on the table with the cursor, a right-click opens a menu, in which the Turn on/off mode shapes dialog
can be activated.
Advanced Step by step tutorial 67

In the window that appears, several filtering condition can be set as shown in the next figure. For example,
the mode shapes that has a modal mass factor less than a certain limit can be turn off (different limit can
be set for each direction). Let us use this function, adjust the criteria of 1% for both directions as shown
below (now it is enough to enter only the limit of x and y).

Mark also the function Reapply after every vibration analysis. If a new calculation is made for any
reason, the set filtering procedure will be performed automatically, but do not forget to check the
achieved total mass factor in the main directions.

Click on the OK button to close the dialog. Now, examine the results and the total modal mass factor in
each direction:

The requirements are satisfied based on the 9 active mode shapes.

Let us look more closely, the number of active shapes could still be reduced. Based on the results, the
first five modes would be sufficient to meet the requirements, the remaining four may be negligible:
68

Let us go back to the previous setting, do not apply this simplification now.

In the tree list on the left, the table called Frequencies is also accessible. This table gives overall infor-
mation about the mode frequencies, period, angular frequency, eigenvalue and error for each mode
shapes.

Close the Table browser. Click on OK to confirm the set filtering criteria.

In the main window, open the list containing the result of mode shapes. Due to the filtering criteria
applies, some of the shapes are shown by pale grey color which also indicates their inactive status.
The active or inactive status of the shapes is also shown on the Status palette.
Advanced Step by step tutorial 69

Setting the necessary filtering criteria, return to the Loads tab to generate the seismic load based on the
results of vibration analysis.

Load cases and Click on the Load cases and load groups icon and create a new Seismic load case.
load groups In the Add panel, click on the icon that symbolizes a response spectrum curve. As a result, the program
creates two ungrouped cases with names SM X, SM Y and a new load group (SEISMIC) is also added to
the list containing SM + and SM – load cases:

Seismic

The program will automatically generate other load cases necessary for the analysis after setting the
seismic parameters. Confirm changes with clicking on the OK button.

The program automatically set the case SM X as current load case and the Seismic load icon becomes
available on the Loads panel (this function is only active if a seismic type load case is selected).
70

Seismic load Click on the Seismic load icon to set earthquake parameters, then the program displays a warning mes-
sage:

Save the model, click on the OK button.

After saving, the following Seismic load parameter window appears:

Adjust the required parameters for our example, let us go through the options:

- at Analysis the type of the vibration analysis can be selected, but only those that have been
already performed: this can be Linear / Nonlinear (see User’s Manual for details of a nonlinear
vibration analysis). This function is now inactive because only one, a linear analysis has been
performed before, so there is no choice.
- next to the Analysis, the Case can be selected (the load case or load combination which was
considered in the vibration analysis). Only one analysis has been made (considering SEISMIC
load combination), therefore it is selected automatically and the combo box is in grey, inactive
state, which means there is no choice.
- under the Case, the Importance factor can be specified (gI). The default value is 1 which equal
to the parameter to be applied, so does not have to change this.
- next to the Importance factor, the displacement behaviour factor has to be set (EN 1998-
1:2008-1 4.3.4 (1)). Adjust this to 1.50, this will be the same as the behaviour factor, see later.

- the necessary response spectrum has to be specified on the Spectrum (horizontal) tab. Under
the Design spectrum several choices are offered:
a) the response spectrum curve can be given by parametric shape based on the regulation
of the standard,
b) Custom spectrum can also be specified,
c) or existing curves can be loaded as well:
Advanced Step by step tutorial 71

Spectral function Applying Custom spectrum, the Spectral function editor has to be used. The spectrum can be
editor… specified by using a formula or complete series of data can be pasted to the cells (e.g. copying
Excel data considering the table format of AxisVM).
The using of the editor is not presented now, for more information please see the relevant
chapter of the User’s Manual.

In the following, let us define a Parametric shape. The parameters that should be specified are
on left side (defining a custom spectrum or loading an existing one, these input fields will be
enabled).

- the reference value of peak ground acceleration is 1,0 m/s2, and the behaviour factor is 1.50
based on the basic data described in the introduction.
If it is justified by the structural concept, different q behaviour factor can be applied in X and
Y directions. In this case, the checkbox in the top left corner must be marked to be enabled the
editing of different parameters. Do not mark this option, because the same factor has to be
used for both direction in our example.

- on the left side the type of the soil (A…E) according to the classification of the standard and
the type of the earthquake response spectrum (Type 1 or Type 2) can be specified. According
to the basic data, spectrum Type 1 and C soil class has to be used. Select the appropriate row
in the list: C Type 1.

Changing the parameters, the program displays the response spectrum curve on the right side
indicating the main characteristic acceleration values (Sd).

Save The main parameters of the standard curve (S, TB, TC, TD, b) are shown under the behaviour
spectrum as… factor. If necessary, these values can also be modified and the result curve can be saved with
clicking on the Save spectrum as… function.

Auxiliary lines On the response spectrum curve, the red auxiliary lines show the period time of the active
mode shapes. They give information about the range of acceleration values which will be used
to generate seismic loads. These auxiliary lines can be turned off clicking on its icon next to
the Save spectrum as… one.

Setting the parameters above, click on the Spectrum (vertical) tab. If it is necessary and it is justified
based on the type of the expected earthquake, the vertical component of the earthquake can also be
72

considered with setting the relevant parameters on this tab. The data input is similar to the previous one.
In our example the vertical component is neglected, let us skip this tab.

Step to the Torsional effect tab and define the necessary data:

Firstly, mark the checkbox (Apply torsional effects) in top left corner.

According to the standard the recommended Accidental eccentricity coefficient is 0.05. It is the default
value, no need to overwrite it.

In the next step, the stories (to which the eccentric loads will be assigned) have to be specified:

If there are existing stories in the model (see earlier), the list is automatically uploaded as shown in the
figure below. If necessary, the generated list can be modified (independently of the Stories that has a
role in displaying parts of the model only): items can be removed (Delete) or new ones can be inserted.
New stories can be added to the list in different ways:
Set the specific height in the field of Z[m] and click on the Add icon to insert stories to the list one by
one. Using the Pick-up function, the height points of the stories has to be marked by the cursor in the
main window. More than one can also be specified at once with this function.

The uploaded stories are correct, do not modify the set.

beszúrt Step to the last tab called Combination methods.

The last tab of the Seismic Load dialog window provides control over the combination of results for
individual modes and individual directions.
Results from individual modes can be combined using either:
Advanced Step by step tutorial 73

the Square Root of Sum of Squares (SRSS),

the Complete Quadratic Combination (CQC) method.
The latter is considered more appropriate if the vibration modes of the structure are not well separated
(i.e. vibration frequencies are close to each other).
Selecting the automatic option lets the program automatically decide if application of the CQC method
is warranted by the vibration results. The program considers modes i and j well separated if the following
condition holds: Tj / Ti < 0.9.
The last option is the default, do not change this in our example.

The value of viscous damping x’ can also be set. The default value is 0.05, no need to change this in our
example.

Results in the two horizontal and the vertical direction can be combined using either of the two com-
monly used combination methods displayed in the dialog window. Use also the default setting as shown
below:

Finish the data input and close the window with clicking on the OK button. After the program shows a
warning message about the lack of the required modal mass factor in Z direction (report shows 0%
because the vertical component of the masses was not considered in our vibration analysis).

Since the vertical (global Z) seismic component is not considered, this message can be ignored. Go on
with clicking on the OK button, then the software generates the load cases for each mode and direction
according to the settings defined by the user.
74

The following naming convention is used for these load cases:

Load cases ending in a 01X, 02X, ....nX, 01Y, 02Y, ....nY, 01Z,
02Z, ....nZ are the equivalent forces in X,
Y or Z direction corresponding to individual mode shapes.
Load cases ending in 01tX, 02tX, ....ntX, 01tY, 02tY, ....ntY
are the torsional moments due to accidental eccentricity in X
or Y direction.

Note: in our example, 9 mode shapes were marked as active.

Considering X and Y directional seismic loads and their tor-
sional effect, the program generates 2x(1+1)=4 cases for
each active mode shape. A total of 9x4=36 load cases have
been generated.

(It was previously mentioned that the first 5 modes shapes

would have been sufficient according to the standard. On this
way, a total of 20 cases would have only been created.)

In the drop-down list of the load cases, choose SM01Y load case and switch on the displaying of loads.
The program displays the concentrated seismic loads on the nodes (inner mesh nodes as well) that have
been assigned to the current load case:

If the nodal load is relatively ‘small’, the value of the concentrated load may seem to be ‘zero’ depending
on the displayed decimals. If necessary, modify the decimals assigned to the unit (Loads/Force) to see
more decimals. The function is available here: Settings/Units and formats/Loads/Force.
Advanced Step by step tutorial 75

Note: if the mesh or a part of it is deleted because of any modification, then the seismic load will also be
cleared.

The other generated load cases are reserved for the results, the following serve for their explanation:

Modal response spectrum analysis uses a combination of linear static analysis results to get the design
seismic effects for the structure under consideration. After the load cases for each vibration mode in each
direction are created, the next step in the MRSA procedure is to run the linear static analyses. The pro-
gram will automatically calculate the effect of each mode in each direction and combine these effects
according to the settings in the Seismic Load dialog window.

After running linear static analyses, there are several seismic load cases available in the list of results
under the Static tab. We use the following naming convention for seismic load cases:

Besides load cases corresponding to individual modes, there are two types of additional results. Load
case names ending in X, Y, or Z contain the combined response from modal results in the X, Y or Z
direction. If accidental eccentricity is considered, there is an additional a or b in the name of load cases
corresponding to horizontal directions. The letters a or b correspond to torsional effects with positive or
negative eccentricity, respectively. Ya for instance is a combination of modal results in the Y direction
considering the effect of positive eccentricity of seismic masses in the X direction.

Combination of load cases in X, Y, Z directions is performed by the program automatically. Such combi-
nation yields a single, unanimous result if there is no accidental eccentricity. If accidental eccentricity is
assumed, the combination of effects from several directions is more ambiguous. There are four basic
cases considered in the program depending on the direction of eccentricity in the X and Y load cases.
Each of the following combinations are performed and stored in a load case with its name ending with
the particular number:

1 = Xa + Ya + Z
2 = Xa + Yb + Z
3 = Xb + Ya + Z
4 = Xb + Yb + Z

MRSA results are absolute values by definition. Therefore, the load cases presented above would contain
only positive values. The least favorable loading scenario is a combination of seismic effects and effects
from other sources such as gravity loading. Creation of this load combination is facilitated in the program
76

by providing two load cases for each seismic load: seismic effects with only positive and seismic effects
with only negative values are identified by a + and a – sign at the end of the name of their load cases.
Note that the absolute values of corresponding responses (internal forces, displacements, etc.) are iden-
tical in the + and – load cases.

Note: displacement results shown in the Static tab are automatically scaled by the qd factor specified in
the Seismic load dialog window.

Table browser Let us check the set earthquake parameters.

Click on the Table Browser icon, on the left side of the window that pops up find and choose the Seismic
parameters:

The table on the right shows the seismic parameters set before.

The model is ready for the seismic analysis.

Change tab to Static. A warning dialog window pops up to save the model before the analysis start.
(Certainly, if there were previous results of static calculations, they will be deleted.)

Click on the Yes button to save the model.

Linear static After click on the Linear static analysis icon, then the calculation starts.
analysis
Advanced Step by step tutorial 77

During the analysis, the program gives information about the actual steps of the calculation, the number
of the equations, the available memory, etc… in the usual way. When the analysis is completed, click on
the OK button to close the window and to return to the main window.

The program shows a warning message about the validity of calculation results because stiffness reduc-
tion has been applied previously for concrete elements.
78

Keep in mind and go on with clicking on the OK button.

Result Considering the above, let us create a seismic combination to inquiring critical values of results. Click on
display the Result display parameters icon, then the following window pops up:
parameters

Here, click on the Critical button below the header, then the combination method of results appears in
the left side of the window. By default, the Critical combinations formula is set to Auto, which contains
all the ULS combinations. In this set the static and seismic type results are also included. Therefore, we
have to separate the seismic results.

Select Custom combination, then the program creates different ULS and SLS combinations to separate
the results.

In persistent and transient design situation, there are two options: ULS and ULS (a,b). Let us use the
default setting - ULS:
Advanced Step by step tutorial 79

In the list of different combinations (being enabled by clicking on the Custom switch) mark the ULS
(Seismic) combination. The latter will only result that this combination mode will be active when querying
results after closing the present window.

Click on OK to close the window. As a result of above, the different combination modes appear in the
combinations list, including the seismic one. The latter may be used to get the critical results (ULS(Seis-
mic)) Critical Min; Max and Min, Max.

The results may be evaluated in the same way as it was presented before in our previous examples, it is
not described in this chapter.

Table browser Click on the Table browser icon and have a look at the table of Seismic sensitivity of stories (this is
only available if the torsional effect has been considered in the analysis).

In the table, the stories are summarized showing their height Z [m], the distance between the stories h
[m] and the following quantities calculated by the program:

θmax – plastic stability index, also known as the interstory drift sensitivity coefficient as per EN 1998-
1:2008 -1 4.4.2.2 (2). Second-order effects (P-Δ effects) need not be taken into account if the θ ≤ 0.1
condition is fulfilled in all storeys. If 0.1 < θ ≤ 0.2, the second-order effects may approximately be taken
80

into account by multiplying the relevant seismic action effects by a factor equal to 1/(1 - θ). The value of
the coefficient θ shall not exceed 0.3. For more details see the relevant chapter in the standard.

Ptot – total gravity load above the storey,

Vtot – total seismic storey shear,

dmax – design interstory displacement calculated as the relative displacement of corresponding storey
centroids,

S – location of the shear center. This is a result of approximate calculation: AxisVM calculates story
shear centers by finding wall sections and using the same method as for thin-walled cross-sections.

Gm – location of the centroid,

M – storey mass,

Imz – moment of inertia at the centroid about the Z axis,

Close the Table browser. The location of the shear center and the centroid is also shown on the finite
element model:

- shear center: S1…n (displayed as red +, with a label Si, where i is the level number)
- centroid: Gm1…n (a cross in a black circle with a label Gmi, where i is the level number).

For better visibility, display only the parts of the slabs:

Advanced Step by step tutorial 81

If the markers of the shear center and the centroid are not visible in the model, it can be switched on in
the Display options (on Labels tab – Story center of gravity, Story shear center):

Additional steps The results of the MRSA can be evaluated in the design modules similar to the simple static results. This
procedure is not described in this example.

If it is justified because of the sensitivity of stories, the second order effect can be considered multiplying
the relevant seismic action effect with fse factor.

In case of dissipative structures with DCM or DCH ductility class, in order to avoid shear failure in RC
columns, beams and ribs, the design values of shear forces are determined in accordance with the ca-
pacity design rule (if seismic load case is included in the selected load combination). Capacity design is
82

only available if the q behaviour factor is greater than 1.50 (choosing Eurocode – RO code, it is available
if q is greater than or equal to 1.50).
Advanced Step by step tutorial 83

Notes
84

Notes